Robot Contact Control Using Flexible Arm Feedback Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for robot control in high-precision contact operations, such as assembly tasks, face challenges in achieving sub-mm accuracy, require complex design changes, and involve high costs due to the need for precise positioning and force control, making them inefficient for fast and reliable performance.
Innovation Solution
A control apparatus for robots with a physically flexible portion that uses machine learning to control the robot's actions based on state observation data, allowing for high-speed operations without the need for complex force control, by integrating a gripper and arm with a flexible portion and employing learning models to determine appropriate actions for contact tasks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If high precision machines and mechanisms are used to reduce uncertainty in position and orientation, then positioning accuracy is improved, but device complexity and design effort increase significantly
Solution Approach 1:
The invention changes the control parameter from position control to impedance control. By modifying the robot's mechanical impedance parameters (stiffness and damping) through control algorithms, the system achieves accurate contact operations without requiring complex mechanical positioning mechanisms. The robot controller adjusts impedance parameters dynamically during contact tasks, allowing the robot to adapt to position uncertainties while maintaining task accuracy.
2Force
If force control is used to control the application of force, then force control accuracy is improved, but control cycle time increases and cost increases
Solution Approach 1:
The invention implements dynamic impedance control where the robot's impedance parameters are adjusted in real-time based on task requirements and contact state. The controller dynamically modifies stiffness and damping characteristics during operation, allowing fast response without sacrificing force control accuracy. This dynamic adaptation enables high-speed control cycles while maintaining appropriate force levels for contact tasks.
3Reliability
If passive operation mechanisms with compliance units are used to absorb errors, then robustness to position errors is improved, but initial positioning accuracy requirements increase and setup time increases
Solution Approach 1:
The invention replaces mechanical compliance units with software-based impedance control. Instead of using physical springs or compliant mechanisms to absorb positioning errors, the robot controller uses algorithms to dynamically adjust impedance parameters. This substitution eliminates the need for complex mechanical compliance mechanisms while achieving the same error absorption effect, and allows for faster response without mechanical inertia limitations.
4Measurement precision
If visual sensors are used to estimate target position and orientation, then positioning capability is improved, but measurement accuracy is insufficient for sub-mm requirements
Solution Approach 1:
The invention implements impedance control with sensory feedback from force/torque sensors and position sensors. During contact operations, the robot controller continuously receives feedback about contact forces and position deviations, then adjusts impedance parameters in real-time to maintain accurate positioning. This feedback-based control compensates for initial positioning uncertainties and achieves sub-mm accuracy without relying solely on visual estimation.
Data Source
Figure 1
Figure 2A~2B
Figure 3
AI summary
Provided is a robot system that can perform operations involving contact at high speed that is cheaper and more general than conventional robot systems. A control apparatus of a robot that is provided with a gripper configured to grip an object and an arm configured to move the gripper, and includes a physically flexible portion provided at at least one of an intermediate position of the gripper, a position between the gripper and the arm, and an intermediate position of the arm, the control apparatus comprising: a state obtaining unit configured to obtain state observation data including flexible related observation data, which is observation data regarding a state of at least one of the flexible portion, a portion of the robot on a side where the object is gripped relative to the flexible portion, and the gripped object; and a controller configured to control the robot so as to output an action to be performed by the robot to perform predetermined work on the object, in response to receiving the state observation data, based on output obtained as a result of inputting the state observation data obtained by the state obtaining unit to a learning model, the learning model being learnt in advance through machine learning and included in the controller.